Reproduce “MiniMMD: Minimization of MMD for Sampling, Inference and Integration” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zonghao Chen
The thesis frames minimum maximum mean discrepancy (MMD) as a shared design problem across numerical integration, distribution sampling, and statistical inference, while proposing separate methods for each. It uses weighted kernel quadrature for nested expectations, regularized flows for transport, and adaptive kernels with preconditioned updates for inference. NKQ reports lower error at equal wall-clock time in three synthetic tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.